Methodology to align business and it policies use case from an it company

METHODOLOGY TO ALIGN BUSINESS AND IT POLICIES:
USE CASE FROM AN IT COMPANY
Christophe Feltus1
, Christophe Incoul1
, Jocelyn Aubert1
, Benjamin Gateau1
, André Adelsbach2
, and Marc Camy2
1
Public Research Centre Henri Tudor, Luxembourg
2
Telindus PSF, Luxembourg
Abstract: Governance of IT is becoming more and more necessary in the current financial economic situation. One declination of
that statement is the definition of corporate and IT policies. To improve that matter, the paper has for objective to propose a
methodology for defining policies that are closer to the business processes, and based on the strict definition of a responsibility model
that clarify all actor’s responsibility. This responsibility model is mainly defined based on the three concepts of capability, the
accountability and the commitment. The methodology is illustrated and validated based on a case study conducted in an IT
company.
Index Terms— Governance, Process model, Organizational model, Responsibility model, Policy Engineering, Business IT
Alignment.
I. INTRODUCTION
ccounting scandals of 2002 and more recently ongoing
market crisis highlight the importance of the Corporate
Governance and by consequence: Governance of IT.
Following those scandals, a lot of laws and standards were
published in order on one hand to guarantee the stability of
the financial sector and, by extension, to all sectors of the
industrial economy and in the other hand, to enhance the
governance all of these public and private companies.
Sarbanes-Oxley Act [1], Basel II [2] and EU Directive 95/46
[3] are some of these laws that aim at providing guarantees
over the company’s accountability. The ISO/EIC 38500:2008
[4] is one standard that provides a framework for effective
governance of IT. One of the main constraints imposed by
these laws and standards is to have responsibilities clearly
established and accepted internally by the collaborators and
externally by the stakeholders as well. Unfortunately, by
depicting the responsibility in a large range of IT oriented
frameworks, we come to the conclusion that no global
consensus over a responsibility model exists. The scope of
our review as targeted organizational models from the realm
of IT security, from access control models such as RBAC [7],
UCON [8] and OrBAC [9], up to framework for ICT
governance like Cobit and its RACI chart [10] or the service
management like ITIL [11]. We have also investigated the
area of requirement engineering, through the analyses of role
engineering methods like [12], [13], [14] and [15] and
through EAM (Enterprise Architecture Model) frameworks
like CIMOSA [16] or Togaf [17]. The importance of the
finding regarding the miss of a common understanding over
responsibility has oriented our research and as a
consequence, we propose in this paper firstly to introduce our
innovative responsibility model that has been elaborated
following the review and based on a global comprehension of
the concepts. This model has already been largely
commented in [5] and [6]. It has been designed to be a
structured representation of the responsibility necessary to
achieve a finite set of activities (like those encompassed in a
process). The three main components of the responsibility
model are Capability, Accountability and Commitment. The
capability describes the quality of having the required
qualities or resources to achieve a task, the accountability
describes the state of being answerable about the
achievement of a task, and the commitment is the
engagement of a stakeholder to fulfil a task and the assurance
he will do it. Hence, the usage of our model may not be
dissociated from the usage of those other models and when
we use them together, the organizational model is enhanced
with responsibility concept and is as a consequence closer to
governance requirements.
Fig. 1. Model aggregation
In Fig. 1, Governance requirements are those dictated by
newly arising laws and standards like the need for more
ethics, more commitment or more accountability. The
engineering of these requirements has been performed in [5]
and [6]. The responsibility model represents the model of
responsibility that has been designed based on these
requirements. The organizational model is the model to be
enhanced with the responsibility components and could be
for example the Cobit framework, the Cimosa framework or
a process based enterprise architecture. That last case has
already been investigated in previous work where a link has
been created between ISO/IEC 15504 and the responsibility
model [18].
Secondly the paper proposes a method to instantiate that
responsibility model based on the enterprise description. This
instantiated model is an intermediary model to be linked to
the organizational model (like a workflow). This method is a
five steps approach starting with a phase of information
collection and closing with a corporate policy. It exists a
plethora of definitions of policy. For the purpose of that
paper, we use the following interpretation that a policy is a
set of roles and responsibilities for a dedicated area of a
A

company or for a field of activity and consequently we decide
to illustrate the paper for policy of access control.
The remainder of the paper is organized as follows: the next
section introduces our innovative responsibility model,
Section III introduces the five steps of our proposed
methodology and illustrates it at the meantime by a real case
study from an IT company. Finally, Section IV concludes and
introduces future works.
II. RESPONSIBILITY MODEL
The cornerstone of the methodology we propose in this
article, and which we will detail in Section III, is based on the
concept of responsibility. The model of responsibility in Fig.
2 aims to be generic enough to be applied to all kinds of
organisations, at each abstraction layer and all domains of the
organisation. In short, the organisation represents a structure
that pursues collective goals. This structure encompasses
employees (users) playing roles and that are responsible to
perform processes’ activities. In this model, the notion of
sequence (the workflow) between the different
responsibilities is not represented. Indeed, these transitions
are already defined in the other organizational models
defining the process model like ISO/IEC 15504.
The notion of responsibility is widely used, but no unique
definition exists. According to the literature, we may
however state that commonly accepted definitions of
responsibility encompass the idea of having the obligation to
ensure that something happens. Our previous work [5] shows
that responsibility can be described as a set of three
additional elements that are Capability, Accountability and
Commitment. The relation between responsibility and the
three other concepts is of the form 0.* to 1. That means that
being responsible involves that the possibility to dispose of
many Capacities, Accountabilities and Commitment.
Fig. 2. Responsibility model
Capability describes the quality of having the required
qualities or resources to achieve a task. For instance, a
strategic capability for a given responsibility could be: “A
resource must know the strategic objectives of the
organisation”. An operational capability could be: “The
coach of the resources must have write access to the HR
software”.
Accountability describes the state of being accountable on the
achievement of a task. For instance, a strategic accountability
for a given responsibility could be: “A project leader must
achieve the financial Key Performance Indicators defined for
the project”. An operational accountability could be: “An IT
administrator must give access rights to specific resources of
the organisation to members of the project team”.
Finally, Commitment is the engagement of a stakeholder to
fulfil a task and the assurance that he will do it. For instance,
a strategic commitment for a responsibility could be:” The
Chief Financial Officer accepts to manage the accounting
department and not commit insider dealing”. An operational
commitment could be: “An employee of the procurement
staff accepts not to use the system for his personal use”.
The consistency between concepts may also be examined
based upon the assumption that the capability needed for
assuming a responsibility corresponds to the accountability of
another responsibility (belonging to another user or role).
Both responsibilities’ components capability and
accountability are strongly linked to each other [5]. An
accountability of a role or a person can permit to deduce
capability of another role or person and conversely a
capability stems from accountability (e.g.: The capability
‘‘The coach of the resources must have write access to the
HR software’’ stems from the accountability ‘‘An IT
administrator must give access rights to specific resources
(HR software) of the organization to the coach’’).
III. METHODOLOGY
The methodology described in this section has for objective
to explain how to define the enterprise IT policies according
to the responsibility model. This methodology is a five steps
approach. To facilitate the understanding, we illustrate each
step with a case study.
Fig. 3. SIM Methodology
3.1 STEP 1. Collect of information
The first step has for objective to define the context and to
collect each component that will be formalized in the policy.

STEP 1 input: Inputs of step 1 are elements collected from
business case studies, business processes, business
procedures, and effective practices in the enterprise.
STEP 1 output: Output of step 1 is a formalized and
structured synthesis of the process in natural language.
STEP 1 actions: The actions performed at this step
encompass a number of activities to collect information about
the process and the responsibility components. These
activities are interviews of the key members of the personnel,
analyses of existing process descriptions, analysis of
enterprise referential like the ISO 9000 quality book.
By these activities, we can summarize by: process
responsibilities as well as their composing elements, like
accountabilities, capabilities, and commitments. The existing
relations between responsibilities and responsibility
components.
To illustrate that methodology, we describe this first step
based on a case study. Telindus Luxembourg SA is an ICT
company within the Belgacom Group, offering its services in
the field of telecommunications and information systems.
Telindus SA is ISO 9001 certified and, as such, formally
defined several processes. For the case study the process of
customer complaints will be analysed.
The customer complaints procedure in Fig. 4 defines the
process of opening, the pursuing and the closing of customer
complaints in order to resolve complaints with a short delay
and, thereby, further improving customer satisfaction.
Complaints are registered in a central complaint database and
assigned to an owner. A central complaint database is used to
track complaints throughout their lifecycle and documenting
actions that have been taken to resolve them, but also the
lessons learned to prevent similar complaints or supporting
their resolution more effectively in the future.
Fig. 4. Contestation Client process workflow
3.2 STEP 2 Graphic diagram
The second step translates the process from natural language
into a graphical representation.
STEP 2 input: Input of Step 2 is the synthesis achieved in
step 1.
STEP 2 output: Output of Step 2 is a graphical representation
of the responsibility framework of the analysed process. It
encompasses a representation of the responsibility and its
components, and the links between components. The Fig.9 is
an example of result obtained with the Telindus Case study.
STEP 2 actions: The actions performed at that step are
composed of three sub-tasks.
Sub-task 1: Definition of each responsibility and transcription
of it using boxes. Each box stands for a responsibility; it
encompasses its accountabilities and its capabilities. An
example identified within the case study is the responsibility
“Creation of Complaint Report” in the contestation process.
Sub-task 2: For each responsibility, an analysis of the
required capabilities is made and is translated through a box
in the corresponding responsibility box. The same operation
is performed for the accountabilities. The “Creation of
Complaint Report” responsibility requires the capability to
receive customer complaints and write access to the central
complaint database. Accountabilities of this responsibility
are, amongst others, to register the complaint in the database.
Sub-task 3: This last sub-task consists in the definition of
links between responsibilities components. Four kinds of
links exist:
- Delegation link constituting the delegation of a
responsibility’s accountability toward another responsibility.
The responsibility of the achievement of the task is
transferred to another responsibility, but the state of being
answerable about the achievement of the task persists for the
delegating responsibility.
Fig. 5. Delegation link
For instance, Fig. 5. represents a delegation of the
accountability #ACC_1 of the responsibility
Responsibility#A towards the responsibility
Responsibility#B. In the case study the accountability
“validation of complaint” of the responsibility “Creation of
Complaint Report” is delegated to the responsibility
“Confirmation/Validation of Complaint”.
- Implication link representing the connection
between the accountability of one responsibility with the
capability of another responsibility. It permits to formulate
that this accountability is needed to provide and guarantee a
capability to another responsibility.
It is important to note that an accountability of a
responsibility may not aim at providing capability to the
same responsibility. Conceptually, this situation would mean
that this responsibility is divisible in two responsibilities.

Fig. 6. Implication link
For instance, Fig. 6. illustrates that the accountability
#ACC_1 of the responsibility Responsibility#A implies the
capability #CAP_1 of the responsibility Responsibility#2. As
an example consider the implication from accountability
“Assure the first level of complaint closure” (“Resolution
Acknowledgement” responsibility) to the capability
“Acknowledgement of the complaint closure”.
- Contribution link highlighting that one
responsibility’s accountability contributes to another
accountability of the same responsibility. Accountabilities
results can be used as input for others accountabilities.
Fig. 7. Contribution link
For instance, Fig. 7. represents that the accountability
#ACC_1 contributes to the accountability #ACC_2 of the
same responsibility Responsibility#A. Considering the case
study (Figure 9), we see that accountability “Register the
complaint in the database” contributes to the accountability
“Assign complaint to the RBL technical or commercial
assistant”, because the complaint can be assigned by the
database when registering it.
- Execution link formalizing that a capability of a
responsibility is necessary to execute an accountability of the
same responsibility.
Fig. 8. Execution link
For instance, Fig. 8. illustrates that the capability #CAP_1 is
needed to achieve the accountability #ACC_1 (written inside
brackets in the capability CAP_1 definition). The capability
#CAP_2 is needed for achieving both accountabilities
#ACC_1 and #ACC_2 (because of the #ALL reference
written inside brackets).. An example from the case study is
that the capability “read access rights in the complaints
database” is needed to “Verify the evolution of the complaint
until closure” (responsibility “Follow-up of the taken
actions”)
3.3 STEP 3 Responsibility’s components links verification
This third step of the methodology is the first refining step. It
aims at analysing the graphical representation of the process
issued from Step 2, depicting and eliminating inconsistencies
from the diagram.
STEP 3 input: Input of Step 3 is the process graphical
representation issued from Step 2.
of the responsibility framework of the analysed process
refined according to the components relationships.
STEP 3 actions: The actions performed at that step are
composed of three sub-tasks.
Sub-task 1: Deep analysis of the capability components for
each responsibility. The main objectives of this analysis are
to detect and solve the problem of unnecessary capabilities.
Capabilities may be unnecessary in the case of useless
capabilities for the achievement of accountabilities of the
same responsibility. This means that they do not have
execution links. To face this inconsistency, it is necessary to
suppress the capability.
Sub-task 2: Deep analysis of the accountability components
for each responsibility. The main objective of this analysis is
to make sure of that all accountabilities are provided and
exist in the model, and to assure that all accountabilities are
necessary. Some accountabilities are not fully justified if:
a. No link exist between the accountability with one or
more capabilities in the process. That means that
there’s no implication link starting from the
accountability
b. No link exist between the accountability and another
responsibility. That means that there’s no delegation
link starting from this accountability
c. The accountability does not contribute to achieve the
outcome of the process.
Sub-task 3: Once accountabilities are verified, it is possible to
check that all capabilities necessary for their achievement
exist.
3.4 STEP 4 Responsibility’s exceptions links verification
This fourth step of the methodology is the second refining
step. As step 3, it aims at analysing the graphical
representation of the relation within the process, in order to
depict inconsistencies and correct the graph to eliminate them
if necessary.
STEP 4 input: Input of Step 4 is the process graphical
representation issued from Step 3.
of the responsibility framework of the analysed process
refined according the relationship between components.

Fig. 9. Contestation Client Process responsibility model
STEP 4 actions: The activity of that step aims at detecting
and correcting conflicts and incoherencies with regard to
responsibility rules dictated by the enterprises, for example:
- Delegation rules. The organisation should define
rules for delegation, which must be complied. Example of
that delegation rules are: if a responsibility is delegated, all
the capabilities necessary for it are also delegated and the
accountability may be kept by the delegator or given to the
delegate but not both at the same time. Some conflict may
exist regarding that rule.
- Separation of duties. Some corporate rules may
impose the separation of duties for some responsibility
components. At this step, a check has to be done, in order to
detect responsibility components that can potentially confer
too much power, in order to prevent frauds or errors. It is
traditionally the case of the accountability to order product
and the accountability to validate the invoice of the product
order. In order to emend such business defects, a
dissemination of responsibility components among multiple
responsibilities has to be done.
- Cardinality constraints. The responsibility graph
needs also to be checked at this step for alignment with
cardinality requirement. E.g.: the number of accountabilities
handled by a same responsibility is sometimes limited in
order to avoid an unjustified increasing working. This
constraint is to be balanced according to the work effort
necessary for achieving each accountability.
3.5 STEP 5 Policy elicitation
This last step of our methodology aims at derogating policies
from the responsibility model.
STEP 5 input: Input of step 5 is the process graphical
representation issued from step 4.
STEP 5 output: Output of step 5 is a set of context dependant
policies.
STEP 5 actions: The activity of that step aims at translating
the responsibility graph in given policy format.
Sub-task 1: Each responsibility is assigned to an
organisational role (such as Project Manager). In other words,
capabilities and accountabilities of each responsibility are
allocated to the roles. Different checks have to be done on
this first instantiated model, in order to detect inconsistencies.
Compliance to rules, checked during the previous step is
again tested. For example, separation of duties (a
responsibility can be protected against abuse of power, but
the combination of two responsibilities for the same role may
enable abuse) or cardinality constraints (individually a load
of work for a responsibility can be supportable, but the

combination of many responsibilities may become
insufferable. This check may include not only the current
process level but also the organisational level).
Sub-task 2: Combining the role instantiated diagram and a set
of roles allocated to organisation stakeholders, the diagram is
instantiated to stakeholders, in which capabilities and
accountabilities of responsibilities are allocated to
stakeholders. Again, checks have to be done, in order to
detect inconsistencies at this step (in the case of stakeholder
having different process involved roles).
Fig. 10 Responsibilities instantiation
Sub-task 3: At this step, the diagram includes all minimum
rights required by all the involved stakeholders for achieving
process (by achieving all accountabilities). Using this
responsibilities distribution, IT policies can be inferred using
capabilities as authorizations, and accountabilities as
obligations. A possible representation of these policies,
described in [18], is the declarative control policy language
XACML. Afterwards, the policy is deployed on the IT
resources via a multi-agent system.
IV. CONCLUSIONS AND FUTURE WORKS
In the current economic context, improving ICT governance
is an important matter. We propose in this paper to improve
that field by introducing our formalization of the
responsibility in an innovative responsibility model. This
model is valuable when it is linked to another existing
organizational model like Cobit or Cimosa and brings to that
organizational model more guarantees regarding corporate
governance requirements. The paper proposes also a
methodology allowing defining, structure and managing the
organization’s responsibilities. To enhance and validate our
work, we have deployed the methodology using the
‘‘Customer Complaints’’ process of Telindus Luxembourg
SA. The case study lead on the one hand to potential
improvements of the customer complaint process, while, on
the other hand, it allowed to validate the methodology, but
also to identify interesting ways of improving and extending
it in future research.
Future works, based on the conclusion of the case study, will
consist on improving the methodology with the addition of a
global iterative refining layer. This layer aims at refining the
responsibility models of the same domain together.
V. ACKNOWLEDGEMENT
The results presented in this paper are a contribution from the
SIM (Secure Identity Management) project and the RED
(Reaction After Detection) project [19].
VI. REFERENCES
[1] P. S.Sarbanes and M. Oxley, “Sarbanes-Oxley Act of 2002”, 2002.
[2] Bank for International Settlements BIS: International Convergence of
Capital Measurement and Capital Standards: Revised Framework –
Comprehensive Version, 2006.
[3] European Union: Directive 95/46/EC of the European Parliament and
of the Council. Official Journal of the European Communities, pp. 28-
31, 1995.
[4] ISO/IEC 15504, “Information Technology – Process assessment”,
(parts 1-5), 2003-2006
[5] Christophe Feltus, André Rifaut, An Ontology for Requirements
Analysis of Managers’ Policies in Financial Institutions, I-ESA2007,
Madeira, Portugal.
[6] André Rifaut, Christophe Feltus, Improving Operational Risk
Management Systems by Formalizing the Basel II Regulation with
Goal Models and the ISO/IEC 15504 Approach, REMO2V’2006,
Luxembourg
[7] D. F. Ferraiolo, R. Sandhu, S. Gavrila, D. R.Kuhn and R.
Chandramouli, Proposed NIST Standard for Role-Based Access
Control, ACM Transactions on Information and System Security, Vol.
4, No. 3, August 2001, Pages 224-274.
[8] R. Sandhu, J. Park, Usage Control: A Vision for Next Generation
Access Control, The Second International Workshop on Mathematical
Methods, Models and Architectures for Computer Networks Security,
2003.
[9] Abou El Kalam, A., El Baida, R., Balbiani, P., Benferhat, S., Cuppens,
F., Deswarte, Y., Miège, A., Saurel, C., Trouessin, G. (2003),
Organization-Based Access Control, IEEE 4th International
Workshop on Policies for Distributed Systems and Networks
(Policy’03), 4-6 juin 2003, Côme, Italie, pp 120-131
[10] Control Objectives for Information and Related Technology (COBIT),
Information Systems Audit and Control Association,
http://www.isaca.org/Template.cfm?Section=COBIT6&Template=/Ta
ggedPage/TaggedPageDisplay.cfm&TPLID=55&ContentID=7981
[11] U.K. Office of Government Commerce, “A Code of Practice for IT
Service Management,” in Service Support, ITIL Managing Services,
Stationery Office, London, United Kingdom (2005), Section 7.8,
http://www.tsoshop.co.uk/bookstore.asp?FO=1159966&Action=Book
&ProductID=0113300158.
[12] Bertino, E., Mileo, A., and Provetti, A. 2005. PDL with Preferences.
IEEE international Workshop on Policies For Distributed Systems and
Networks, Policy 2005 – Vol. 00, IEEE Computer Society,
Washington, DC, 213-222.
[13] Yu, E. S. and Liu, L. 2001. Modelling Trust for System Design Using
the i* Strategic Actors Framework. Workshop on Deception, Fraud,
and Trust in Agent Societies Held During the Autonomous, Eds.
Lecture Notes In Computer Science, vol. 2246. Springer-Verlag,
London, 175-194.
[14] A. Antón, Goal-Based Requirements Analysi,. Second ICRE’96,
Colorado Springs, USA, 1996.
[15] Robert Crook, Darrel Ince, Bashar Nuseibeh, Towards an Analytical
Role Modelling Framework for Security Requirements, Security
Requirements Group, Departement of Computing, The Open
University, Walton Hall, Milton Keynes, MK7 6AA, UK.
[16] Vernadat F. B., Enterprise Modelling and Integration, Chapman &
Hall, London (1995), ISBN 0-412-60550-3
[17] Togaf (2007), The Open Group Architecture Framework (TOGAF
8.1.1 'The Book'), 2007 Edition , Van Haren Publishing.
[18] Gateau, B., Feltus, C., Aubert J., Incoul, C. (2008), An Agent-based
Framework for Identity Management: The Unsuspected Relation with
ISO/IEC 15504, RCIS 2008, Morocco.
[19] http://projects.celtic-initiative.org/red/?Dissemination:Publications
Manuscript received September 30, 2008. Corresponding author: C. Feltus
(e-mail: christophe.feltus@tudor.lu)

Methodology to align business and it policies use case from an it company

Related slideshows

Recommended for you

More Related Content

What's hot

What's hot (17)

Viewers also liked

Viewers also liked (14)

Similar to Methodology to align business and it policies use case from an it company

Similar to Methodology to align business and it policies use case from an it company (20)

More from Luxembourg Institute of Science and Technology

More from Luxembourg Institute of Science and Technology (13)

Recently uploaded

Recently uploaded (20)

Methodology to align business and it policies use case from an it company